- Matching
execute
parameters behaviour to those of psycopg2. (issue 136).
- Doc fixes (issue 131). Thanks to gward.
- Makefile fix (issue 132). Thanks to bitwolaiye.
- Catching all syncrhonous exceptions (issue 134). Thanks to m-messiah.
- Catchin
IOError
s in IOLoop handlers (issue 127).
- Fixed serious flaw with connection retrials. More details.
- Fixed ping to handle failures properly (issue 115).
- NOTE: gcc is now required to run tests - we supply built-in version of tcproxy for connection failure simulation.
- Fixed JSON/HSTORE support with named cursors (issue 112). Thanks to helminster.
- Auto shrink support. Thanks to John Chumnanvech.
- Full rewrite using using Futures
- NOTE: The new API is similar but not backwards compatible. Make sure to read documentation first.
- Aadded register_json
- Docs: fix typos, spelling, grammatical errors; improve unclear wording
- Removed support for psycopg2ct
- Catching ALL types of early error. Fixes issue 79.
- Tornado 4.0 compatablity: backported old
Task
class for Tornado 4.0 compatablity.
- Fixed hstore.
- Fixed a minor Python 3.2 issue.
Fixes:
Connection.transaction
does not break when passed SQL strings are ofunicode
type
New features:
- Transparent automatic reconnects if database disappears and comes back.
- Session init commands (
setsession
). - Dynamic pool size stretching. New connections will be opened under load up-to predefined limit.
- API for manual connection management with
getconn
/putconn
. Useful for server-side cursors. - A lot of internal improvements and cleanup.
Fixes:
- Connections are managed explicitly - eliminates transaction problems reported.
connection_factory
(andcurosr_factor
) arguments handled properly byPool
.
- Fix code example in documentation. By matheuspatury in pull request 46
- Tested on CPython 2.6, 2.7, 3.2, 3.3 and PyPy with Psycopg2, psycopg2ct and psycopg2cffi.
- Add and remove a database connection to and from the IOLoop for each operation. See pull request 38 and commits 189323211b and 92940db0a0 for more information.
- Replaced dynamic connection pool with a static one.
- Add support for hstore.
This is a beta release. It means that the code has not been tested thoroughly yet. This first beta release is meant to provide all the functionality of the previous version plus a few additions.
- Most of the code has been rewritten.
- The mogrify method has been added.
- Added support for transactions.
- The query chain and batch have been removed, because
tornado.gen
can be used instead. - Error reporting has bee improved by passing the raised exception to the callback. A callback accepts two arguments: the cursor and the error.
Op
,WaitOp
andWaitAllOps
inmomoko.utils
are wrappers for classes intornado.gen
which raise the error again when one occurs. And the user can capture the exception in the request handler.- A complete set of tests has been added in the
momoko
module:momoko.tests
. These can be run withpython setup.py test
.
- Removed all Adisp related code.
- Refactored connection pool and connection polling.
- Just pass all unspecified arguments to
BlockingPool
andAsyncPool
. Soconnection_factory
can be used again.
- Reorganized classes and files.
- Renamed
momoko.Client
tomomoko.AsyncClient
. - Renamed
momoko.Pool
tomomoko.AsyncPool
. - Added a client and pool for blocking connections,
momoko.BlockingClient
andmomoko.BlockingPool
. - Added
PoolError
to the import list in__init__.py
. - Added an example that uses Tornado's gen module and Swift.
- Callbacks are now optional for
AsyncClient
. AsyncPool
andPoller
now accept a ioloop argument. [fzzbt]- Unit tests have been added. [fzzbt]
- Renamed
momoko.Momoko
tomomoko.Client
. - Programming in blocking-style is now possible with
AdispClient
. - Support for Python 3 has been added.
- The batch and chain fucntion now accepts different arguments. See the documentation for details.
- Removed
executemany
fromMomoko
, because it can not be used in asynchronous mode. - Added a wrapper class,
Momoko
, forPool
,BatchQuery
andQueryChain
. - Added the
QueryChain
class for executing a chain of queries (and callables) in a certain order. - Added the
BatchQuery
class for executing batches of queries at the same time. - Improved
Pool._clean_pool
. It threw anIndexError
when more than one connection needed to be closed.
- Initial release.